TASVideos

Tool-assisted game movies
When human skills are just not enough

Submission #4367: ForgoneMoose's PSX Castlevania: Symphony of the Night "All Relics & Bosses" in 36:58.32

Console: Sony PlayStation
Game name: Castlevania: Symphony of the Night
Game version: USA
ROM filename: sotn.iso
Branch: All Relics & Bosses
Emulator: PSXjin v2.0.2
Movie length: 36:58.32
FrameCount: 133099
Re-record count: 151764
Author's real name: Dan W.
Author's nickname: ForgoneMoose
Submitter: ForgoneMoose
Submitted at: 2014-07-16 00:37:29
Text last edited at: 2014-07-26 18:28:06
Text last edited by: Guga
Download: Download (32920 bytes)
Status: published
Click to view the actual publication
Submission instructions
Discuss this submission (also rating / voting)
List all submissions by this submitter
List pages on this site that refer to this submission
View submission text history
Back to the submission list
ALERTS POSSIBLY COMPROMISING MOVIE INTEGRITY:

Preinitialized memory card 1

Is empty

Preinitialized memory card 2

Block 1
  Name:         'CASTLEVANIA-1 X-X!V??Q 33%' Filename:     'DRAX00'
  Product Code: 'SLUS-00067'          Territory:    American
  Blocks Used:  1                     Bytes Used:   8192
  Status:       OK                    Icon:         3 frames
  Link Type:    First                 Next Block:   None
Author's comments and explanations:
That day, probably in 1956 YEARING. As the destruction of the world just does not know itself, through the entire CHAOS best life, DO physics who can be expected to prevent it alive in each level and free. Attack bitches OUT recommend social workers, a man just now rising human nature mathematics strike day! Death will come soon to the longing for the robot to be busy, it is sufficient if the only man named Thomas

Beyond, who have lost their land, not tied together in this mode is actually quite an enemy, they did not realize it was so harmless, maybe you become tame hurt themselves with the enemy, so much in fact, there is no comparison sorry, I said ? Find out on the next episode of 24

Game Objectives

  • Emulator used: PSXjin v2.0.2
  • 100% completion (close enough)
  • Abuses programming errors
  • Takes damage to save time
  • Manipulates luck
  • Colors the hell out of a vampire



(Link to video)

About Goal Choice

Although this movie has a notably different goal from the previous 100% completion TAS, I absolutely believe that this should be accepted as an obsoletion to it. With the addition of ACE in this game, the use of save corruption is questionable for a 100% category and using it to reset the map like the old TAS does seems very arbitrary. There's no way one could make a TAS that gets the 196% map needed for the best ending without that glitch and still have it be faster or as entertaining as the old TAS, so it is therefore not possible to make a satisfactory straight obsoletion to the old TAS. On the other end of the argument for save corruption, what would be the point in a goal that makes the movie exactly the same as the ACE TAS but with a slightly longer menuing section? How does that show off the whole game? There's just no good way to do any goal wherein save corruption is such a large factor.

Not only that, but I believe that the goal of map completion was a bad goal to choose even if map corruption is allowed. As you can see in this movie, the 20+ minutes shaved off the previous TAS didn't come just from improved movement and glitches. There are just these large periods where arukAdo had to stop the flow of the TAS and just sit there and move around large, empty areas for no reason besides adding map squares, and its really jarring and simply unentertaining to watch when we just want him to get back to speeding through the game, like you'd, you know, expect a TAS to do. He spent so much time just backtracking and doing the same few OOB glitches in the first castle that it just feels like a slog watching it.

Not to mention, there is absolutely a speed/entertainment tradeoff deal in that goal; the fastest map collection would just be doing the same OOB repeatedly and map corrupting, so his route was mostly based around maximizing entertainment. That in mind, it makes the goal feel like a copout for poor routing and movement, almost - anything that looks off is just showing us off another map square, after all! Not that I disagree with the choice to make such tradeoffs, I just don't see how a goal based almost entirely around entertainment over speed is a good idea for what's supposed to be a SPEED run.

Most importantly, there's the fact that this is simply an objectively better made TAS than arukAdo's best ending TAS. It was his first foray into working on this game entirely alone, and it very much shows. The optimization of movement and combat is downright horrendous at some parts, even visible to those not familiar with the game even, and I don't think that this level of quality should be kept on the site when you're being offered an alternative TAS that is visibly better. arukAdo himself acknowledged this, expressing great interest in making a new 100% completion TAS, saying that, "the site deserved an update to it."

As this TAS actually ended up being very similar to the best ending TAS but stripped of the stops to gather map%, I don't see why this can't obsolete it. If anyone doesn't believe my claims of it being significantly better optimized, I can make video comparisons showing any number of instances of much better movement and such, but I think it should be apparent if you just compare them directly.


For the new goal, it was something that I actually spent a long time thinking about, and I didn't even have the goal set in stone until I was actually at Form of Mist. I realized that although the bosses in this game aren't really threatening in casual play in comparison to other all bosses runs, I could make something entertaining just by showing off exactly how pathetic all of them are - creating a sort of art out of destroying every single one of them mere frames after their AI activates.

I still thought that there could be a better goal, as there are obviously some areas went unvisited (of course, the old TAS didn't visit all the areas either). What goal would take me to the Forbidden Library? The Castle Center? The best idea me or arukAdo could come up with was one of every item in the game, but after debating with myself I realized this would get quite repetitive quite fast, especially since I wouldn't even be using almost any of the items I got, instead just using only the same things I used in this submission.

There just simply isn't any logical better goal for going everywhere in the game, so I'm afraid that since map completion is out of the question, we're stuck with all bosses. I would have preferred a category that showed off more of the glitchiness the best ending TAS had, but most of its nonsense only served to collect map% anyway.

I decided what to do about Richter by simply asking in various places related to the game and going with a majority vote. It ended up being mostly split between saving him and using a glitch to kill him without getting the bad ending. As skipping him as console speed runs do in all bosses runs was completely out of the question in regards to the polls, and I don't believe that items/achievements that mandate glitching to get should be counted in the goals of 100% runs period, I decided to just settle for saving him. If anyone has a problem with that, take it up with the rest of the internet, not me.

Stuff About the Run

A big improvement to the first few minutes was ZX497 and I's discovery of the Toadstool damage boost onto the platform leading to Olrox's quarters. As I absolutely need to get mist, bat, and the stuff from the shop no matter what route I take, this is essentially a more direct route than the previous route of getting the early library card in Marble Gallery then going all the way back around to the clock room again. For comparison, I have bat and all the items from the shop by the time the old TAS is still in Olrox's Quarters. It's difficult to compare the two TASes because the old one stops to gather large amounts of map% often, but sometimes it is relatively practical, like in that case.

The discovery of the ability to override level up and relic get animations with summon spirit was something I actually had to start the TAS over halfway through for. It saves a second each time, and that really adds up over the course of a ~40 minute run, to say the least. Unfortunately, most relics in the second castle are too far up in the air to override like this.

The all bosses goal was a double edged sword over the course of the game. On one hand, movement is crisper because I have much more MP than the best ending TAS, but on the other hand, the higher max HP the level ups give me makes big tossing places difficult or impossible around the time I get to the second castle. As I couldn't find spots where big toss saved large time in the second castle, I feel that it was definitely a trade in my favor.

There's a lot to say about the extended movement tech over the any% route here. Gravity boots basically let me make boomerang jumps without using boomerangs, and they move faster horizontally. They can only be canceled by transforming, and hitting a ceiling with them forces you to stop and wait there for a couple seconds like with big toss. The bat's Wing Smash is slower than big tossing, but can be angled and does not have the long stop at the end, plus I don't need to heal between each use. The Wolf Charge tops out at significantly faster than Wing Smash, but it takes a while for it to break Wing Smash speed, and is only useful over very large stretches. My ability to do stuff such as Wing Smash > untransform, gravity boots > cancel into bat is limited by a timer of about half a second before you can transform after untransforming, so movement seems awkward at times because I have to deal with that.

Weapons of Murder

  • Alucard Sword is the starting weapon, and the best sword that can reasonably be obtained in the first castle far and away, but it obviously is weak in comparison to the nigh-instant death that the later weapons cause. It takes a notable penalty in damage output because my STR is so low due to using luck mode. It sees occasional use later in the game because swords use a different invincibility timer from Flame Stars.

  • Flame Star kills have multiple methods of optimization, depending on the HP of the boss you're trying to kill. One one hand, you want to keep your use of stars in groups of 3, as you can only have 3 at a time per hand onscreen at once before you have to wait a moment to start throwing again, so either 3 or 6 star kills are a priority. However, you can only throw 4 stars into a boss in quick succession before you have to wait out a brief invincibility timer, so that is yet another consideration. Divekicks and spells use the same invincibility timer as stars, so I can't really do much of anything else while waiting for the timer to go down.

  • Javelins are weaker than Flame Stars, but both connect with targets faster after being thrown and shotgun through whatever it hits, making it a better weapon for hitting multiple targets or fat targets than flame stars. It is however limited by only being able to throw 2 per hand, however. They travel in an upward arc when thrown, letting me hit enemies somewhat above Alucard that flame stars cannot.

  • Alucard Shield spell is the most powerful weapon throughout the run, but it has a long casting animation before it can be used. Finding any way to override the animation or use it for screen pause glitches saved a lot of time. The spell stays active until you run out of MP, hold the shield up for a total of around 68 seconds, or equip an item over the shield, so optimization was made around keeping the shield equipped as much as possible. The shield hits for a cap of 255 damage 10 times a second - contrary to what some have said, the 255 is subject to enemy's stats and such.

  • Power of Sire is the last of my weapons. I don't use it on Galamoth because he strongly resists it, but it really pays for itself against the final bosses, but mostly for letting me kill The Creature without having to recast the Alucard Shield spell. It does one hit immediately, but has a pause before the rapid fire hits after that first hit, so I use it early against the weak Shaft. It does less damage than it did in the previous TAS, but this is because the old TAS had the Ring of Ares to augment its attack power, which I don't have convenient access to. It's still quite powerful though, obviously.

  • Critical hits are interesting in this game. They usually use a fairly standard formula of adding damage to hits if they occur on certain outputs of the RNG, but this added number can vary anywhere from normal damage+1 to normal damage + double your STR, without any real pattern. However, if your LCK is higher than your STR, the formula is normal damage + a random amount of your LCK - considering I have 119 total LCK right off the bat, I can effectively do double damage per hit in the first castle. Different weapons have different chances to critical hit; Alucard sword has a moderate chance, Flame Stars have a very high chance, and Javelins notably have a very low chance. I have not seen numbers on this effect documented anywhere, I found this through pure personal experience making this TAS.

Finally

More in-depth commentary can be found in the commentary track, and I've explained a lot about minor physics oddities and movement tech in previous submissions as well as the resources page for this game.

I would like to thank ZX497 for helping with some things and basically proofreading the TAS for me, and Desquall for coming up with a number of the strats I used in the first castle, as well as Sockfolder for helping optimize the toadstool damage boost nonsense, as well as helping with various other miscellaneous things.


Mothrayas: Judging.

Mothrayas: It seems the audience agrees that this run should obsolete [1592] PSX Castlevania: Symphony of the Night (USA) "all relics, best ending" by arukAdo in 57:20.6, as this run has a more preferable goal choice and is better optimized. As such, accepting as an improvement to that publication.

Guga: Processing...


Similar submissions (by title and categories where applicable):